The story begins with two stickleback species. Freshwater threespine sticklebacks, Gasterosteus aculeatus, live permanently in the rivers of the region, while the Japan Sea stickleback, Gasterosteus nipponicus, is a marine or sea-run migrant. The two lineages diverged roughly 680,000 years ago and have evolved strong but incomplete reproductive isolation. The tsunami appears to have carried marine G. nipponicus inland from the Pacific Ocean, while the backwash swept freshwater G. aculeatus down from upstream reaches, dumping both species into the same newly formed ponds. By 2012, researchers sampling the ponds found a population teeming with hybrids, something that had never been documented in Otsuchi before the disaster.

What happened next is the heart of the new study, published in Nature Ecology & Evolution. Using restriction-site associated DNA sequencing across nine years of sampling, the team tracked the fate of G. nipponicus ancestry generation by generation. In 2012, 38 percent of sampled fish were hybrids, most of them first-generation backcrosses toward G. aculeatus. Yet by 2020, within roughly ten generations, the population had become almost entirely pure G. aculeatus. The marine species’ genome had been swept out of the population with remarkable speed, a process the researchers describe as genome-wide purging.

Could this rapid loss of foreign alleles simply be a statistical accident of small population size? The researchers tested that possibility rigorously. A mark–recapture survey estimated the census population at 22,485 individuals, and stochastic simulations based on the Wright–Fisher model were run at effective population sizes corresponding to 100 percent, 10 percent and 1 percent of that figure. Starting from the observed allele frequencies at 667 species-diagnostic loci, the simulations showed that at most loci—98.7 percent, 97.8 percent and 85.3 percent respectively—the real population purged G. nipponicus alleles faster than drift alone could ever explain. Selection, not chance, was doing the work.

The next question was where that selection was strongest. The team already knew from previous work that the two species are kept apart by multiple barriers: ecological selection against immigrants and hybrids, differences in habitat choice, sexual isolation and hybrid male sterility. Quantitative trait locus mapping had previously tied sexual isolation and hybrid male sterility to chromosomes 9 and 19, which function as the neo-X and ancestral-X chromosomes in G. nipponicus, a system that arose through a Y–autosome fusion. Strikingly, the new time-series data showed that G. nipponicus alleles at these major reproductive isolation loci vanished within the first few generations, faster than the genomic background.

To understand why, the researchers ran field experiments in the tsunami-created habitat itself. In outdoor net enclosures, pure G. nipponicus suffered dramatically higher mortality than pure G. aculeatus, with backcross hybrids showing intermediate survival. By comparing allele frequencies between the backcross fish released into the enclosure and the survivors six months later, the team identified five candidate loci on chromosomes 10, 15, 17 and 20 and the ancestral X chromosome where marine alleles reduced survival. Interestingly, a well-known freshwater adaptation—an extra copy of the fatty acid desaturase gene Fads2 on chromosome 12—played no role here, likely because prey organisms washed in from the sea kept the ponds well supplied with docosahexaenoic acid.

Habitat choice added a second, equally elegant mechanism. Because the tsunami ponds connect to the estuary only through small tunnels, any fish carrying marine alleles that triggered seaward migration would simply leave, never to return. The researchers built an eight-chambered experimental stream system in the field and found that G. nipponicus migrated seaward far more often than freshwater G. aculeatus, with backcross hybrids again intermediate. Quantitative trait locus mapping revealed significant loci for seaward migration on chromosome 10 and the ancestral X chromosome—regions that overlapped with the survival loci. This clustering of barriers for multiple isolating mechanisms on the same chromosomes, including a large-X effect, means a handful of genomic regions act as a disproportionately powerful barrier to gene flow.

But major loci alone could not explain the full story. Once the deleterious alleles at major barrier loci are purged in the first few generations, no further selection should act on neutral introgressed DNA elsewhere in the genome. Individual-based simulations confirmed this: models with only major reproductive isolation loci, or with drift alone, failed to reproduce both the observed admixture pattern and the subsequent genome-wide purging. The pattern only emerged when the team added many pairs of type 2 Dobzhansky–Muller incompatibilities—negative epistatic interactions between a dominant allele of one species and a recessive allele of the other—scattered across the genome. These incompatibilities do not harm first-generation hybrids but inflict fitness costs on later generations, when alleles become homozygous. In other words, a dense genome-wide network of weak incompatibilities, combined with a few major barriers, appears to be what prevents species fusion.

The pattern repeated itself elsewhere. The team located two other tsunami-struck stickleback habitats, in Iwaizumi and Kuji, roughly 55 and 90 kilometers north of Otsuchi. Both contained recent hybrids with predominantly G. aculeatus ancestry, and the frequencies of G. nipponicus-derived alleles across the genome correlated significantly between habitats, suggesting that the same genomic regions were purged first in each independently founded population. Differences in how much marine ancestry remained appeared to depend on the initial ancestry composition, with simulations showing that higher initial frequencies of G. nipponicus led to greater retention of its alleles. Elevated marine ancestry on chromosomes 12 and 13 in Iwaizumi may reflect heterozygote advantage or adaptive introgression, though the current data cannot rule out drift.

The implications reach well beyond Japanese ponds. The divergence between these two species sits near a proposed transitional threshold of speciation, and the study provides direct evidence that species identity can persist even after extensive introgression, provided enough genetic incompatibilities have accumulated. It also echoes a lesson Darwin drew from an earthquake in Chile aboard the Beagle: catastrophic geological events can open rare windows onto fundamental evolutionary processes. As climate change intensifies natural disasters and human activity increasingly forces species into contact, predicting whether hybridization leads to fusion, extinction or renewed isolation becomes ever more urgent. This nine-year experiment, launched by a tragedy, shows that the genome keeps its own defenses—and that watching them work is now possible in real time.

Subject of Research: Real-time genomic purging of introgressed alleles in a tsunami-induced hybrid stickleback population

Article Title: Rapid genome-wide purging following tsunami-induced hybridization in a stickleback population

Article References: Hosoki, T. K., Mori, S., Kagawa, K., Nishida, S., Kume, M., Nagano, A. J., Kanbe, H., Kakioka, R., Nakamoto, K., Iino, Y., Kodama, M., Oba, S., Seong, T., Kabeya, N., Ishikawa, A., Yoshida, K., Yamasaki, Y. Y., & Kitano, J. (2026). Rapid genome-wide purging following tsunami-induced hybridization in a stickleback population. Nature Ecology & Evolution. https://doi.org/10.1038/s41559-026-03184-1

Image Credits: AI Generated

DOI: 10.1038/s41559-026-03184-1

Keywords: stickleback, hybridization, speciation, genomics, tsunami, introgression, reproductive isolation, Dobzhansky-Muller incompatibilities, purging, evolutionary biology, population genetics, Japan

The review, led by joint first authors Robin Quagebeur and Emma De Schuyteneer, followed the PRISMA-ScR reporting guidelines and searched eleven databases for studies of individuals diagnosed with anorexia nervosa or atypical anorexia nervosa. To be included, a study had to examine negative affect, anxiety, or stress and report outcomes connected to eating-disorder symptoms. The team then sorted the resulting literature into four methodological families: retrospective self-report studies, which ask people to recall past experiences; qualitative studies, which capture lived experience in interviews and narratives; experimental studies, which deliberately provoke emotions under controlled conditions; and daily life studies, which use techniques such as ecological momentary assessment to sample feelings and behaviours in real time as people go about their routines. This taxonomy matters because each method answers a slightly different question, and the review shows that the answers do not always converge.

The experimental strand of the literature provides some of the clearest signals. When researchers induce negative emotions in the laboratory, for example by showing distressing films or asking participants to recall painful memories, people with anorexia reliably report intensification of disorder-specific cognitions: thoughts about weight, shape, dietary rules, and the perceived consequences of eating. In other words, the emotional state does not merely coexist with the illness; it appears to amplify the cognitive machinery that sustains it. This aligns with influential theoretical models of eating disorders, including the affect regulation model and the escape from self-awareness account, which propose that disordered eating behaviours function as maladaptive strategies for managing unbearable emotional states. The experimental evidence reviewed here gives those models a firmer empirical footing, at least for the cognitive component of the disorder.

Daily life studies, which capture behaviour where it actually happens, tell a more differentiated story. Across experience-sampling research, momentary negative affect is consistently associated with binge eating and with purging, the behaviours most directly tied to emotional escape and relief. Yet when the outcome is dietary restriction, the hallmark behaviour of the restrictive subtype of anorexia, the associations are mixed, with some studies finding that negative emotions precede restriction and others finding no reliable link or even the opposite pattern. The same inconsistency applies to physical activity, which for many patients becomes a compulsive means of burning calories. This heterogeneity suggests that restriction and compulsive exercise may not serve the same emotional functions across individuals, and that subtype differences, between the binge-purge and restrictive presentations of the illness, may be critical to understanding when emotion actually drives behaviour.

Anxiety emerges in the review as a particularly mealtime-bound phenomenon. Studies that measured anxiety in the moments surrounding eating found that it was often related to caloric intake, with anticipatory anxiety about eating, body concerns, and the social context of meals shaping how much food people consumed. This is a clinically intuitive but empirically important observation: for many individuals with anorexia, the plate itself is a threat cue, and the anxiety it provokes in the moment may directly determine whether a meal is completed, reduced, or avoided altogether. Interventions that target mealtime anxiety, such as supported meals and exposure-based approaches, therefore have a clear mechanistic rationale, and the review’s findings suggest that measuring anxiety at the moment of eating could help clinicians identify which patients need the most intensive support around food.

Stress, the third pillar of the review, proves to be the most methodologically tangled construct. Most studies relied on general stress paradigms, most prominently the Trier Social Stress Test, a standardized laboratory challenge in which participants deliver a speech and perform mental arithmetic before an evaluative panel. The effects of such challenges on eating-disorder behaviours were mixed, and the review highlights a striking divergence between subjective and physiological stress responses: what people reported feeling did not always match what their bodies were doing. Physiological stress is typically indexed through the hypothalamic-pituitary-adrenal axis, measured via the stress hormone cortisol, and through the sympathetic-adrenomedullary system and broader autonomic nervous system activity. Because self-reported stress and biological stress can come apart, studies that rely on only one channel may be systematically missing part of the picture, and the review argues that this dissociation is itself a key finding deserving of targeted research.

One of the review’s most sobering conclusions is how rarely the three constructs have been compared head to head. Negative affect, anxiety, and stress are conceptually overlapping but distinct: negative affect is a broad valence state, anxiety is a specific fear-oriented emotion often with anticipatory character, and stress encompasses both a psychological appraisal and a physiological response. Yet few studies measured more than one of these within the same design, which makes it impossible to say which construct is most tightly coupled to symptoms or whether they act through shared or separate pathways. The authors argue that future work should adopt multi-modal assessments, combining momentary self-reports with physiological monitoring, potentially including wearable sensors, so that the relative contributions of each affective state can be disentangled rather than assumed.

The methodological inconsistencies documented in the review are not trivial. Studies varied in how they conceptualized negative affect, anxiety, and stress, in the instruments they used, in the timing of their measurements, and in the symptom outcomes they reported. Retrospective questionnaires, for instance, are vulnerable to memory bias and to the way eating disorders distort self-perception, whereas ecological momentary assessment reduces recall demands but introduces its own burdens and reactivity. Experimental paradigms offer control but sacrifice ecological validity. The review’s framework, by making these differences explicit, gives the field a map of where its evidence is strong and where it is fragile, and it provides a template for how future studies could be designed to be comparable with one another rather than isolated data points.

For clinicians, the practical message is that emotional states are not background noise in anorexia nervosa; they are moment-to-moment modulators of symptoms. The strongest and most consistent associations involve the emotional amplification of disorder-specific thoughts and the triggering of binge-purge behaviour, while restriction and compulsive exercise follow more idiosyncratic patterns that may require individualized formulation. Timing interventions to moments of peak negative affect, teaching patients to recognize the emotional antecedents of their urges, and addressing mealtime anxiety directly all emerge as evidence-supported priorities. For researchers, the message is equally clear: the next leap in understanding will come from studies that measure mind and body together, in the same moments, in the same individuals, and that directly compare negative affect, anxiety, and stress rather than treating them as interchangeable labels for the same phenomenon.

As a scoping review, the study maps the terrain rather than pooling effect sizes, and the authors are careful about the limits of their conclusions. The translation from feeling to behaviour, they conclude, varies by context and by illness subtype, and the field’s understanding of the interplay between affective states and symptoms in anorexia nervosa will only advance with stronger methodological consistency. But the central insight stands out with unusual clarity for a body of literature this heterogeneous: in the moments when negative emotions, anxiety, and stress rise, the cognitive and behavioural symptoms of anorexia nervosa are close behind. Capturing those moments, in the laboratory and in daily life, may be the key to predicting, and ultimately interrupting, the cycles that keep the illness alive.

Subject of Research: Momentary associations of negative affect, anxiety, and stress with symptoms in anorexia nervosa

Article Title: The momentary associations of negative affect, anxiety and stress with symptoms in individuals with anorexia nervosa: a scoping review

Article References: Quagebeur, R., De Schuyteneer, E., Vergaelen, E., Presseller, E. K., & Vrieze, E. (2026). The momentary associations of negative affect, anxiety and stress with symptoms in individuals with anorexia nervosa: a scoping review. Journal of Eating Disorders. https://doi.org/10.1186/s40337-026-01744-1

Image Credits: AI Generated

DOI: 10.1186/s40337-026-01744-1

Keywords: anorexia nervosa, negative affect, anxiety, stress, eating disorders, ecological momentary assessment, binge eating, purging, dietary restriction, mealtime anxiety, scoping review, physiological stress response